Is Promax PX able to derive future demand forecasts from past business data using a library of forecasting algorithms? By utilizing a number of forecast methods the system must be able to calculate best fit based on MAD/SAD or user defined parameters. Best fit parameters may be defined differently by group.
Promax PX uses a library of forecasting techniques that derive statistical forecasts for promotional uplift and baseline sales. The choice of technique is determined by the source of data which can vary from POS Scan data from AC Nielsen or Synovate/Aztec to warehouse withdrawal information or ex-factory sales. Promax PX allows the user to define forecast algorithms and has a library of algorithmic forms on which it employs a tournament system to recommend the best model based on “goodness of fit”. The recommendation can be ranked by a number of measures including Standard Error, Correlation Coefficient or residual.
Promax PX uses non-linear regression as the foundation of the modelling approach and Promax has devised a flexible and specific approach that is tailored to the needs of both the forecaster and the trade promotions planner.
The modelling of scan data includes imbedded technology to filter the historical data into its key elements. The sales pattern is made up of such elements as trend, seasonality, baseline and promotions. These elements are affected by selling price, display, marketing vehicle, advertising and peripheral events like competitor promotions, community calendars (public holidays, Easter, to name some examples. The causal components can in themselves have effect on the baseline or the promotional uplift and the modelling engine needs to be able to distinguish between these.
At Promax we see this as essential, modelling through a tournament method which tests many algorithmic forms to determine the “best” one that describes the pattern of sales. With this approach each product/account combination could have a different model. It gives the user the ability to use models developed by others like your syndicated data providers (ACNielsen, IRI or Synovate/Aztec) leveraging your previously made analytics investment.
The solution design means that Sales users can eaily select / enter the model peramiters when developing promotional plans – which seamlessly flows into the development of a sales forecast thus truly ensuring a ‘one number’ approach to the business
With Promax PX are you able to choose from different forecasting units, including sales value, cost value, weight, volume and stock unit when maintain forecasts?
Promax PX is able to hold Forecasts and Sales history at both the Cases and Units as a base level. Promax PX holds information on:
- Units Per Case
- Standard Cost / COGS
- List Price
- Case Weight / Litre (Gross & Net)
Using this information combined with the cost of trade spend related activity Promax PX is able to give both a Gross Sales & Net Sales Forecast. The Promax PX database holds both the forecast and actual sales information allowing a closing a position (Outlook) for the financial period to be calculated. Promax AnalytX BI (OLAP Data Cube) allows any number of calculations to be performed using these measures providing limitless configurability and customisation for review & forecast reporting.
Will Promax PX allow the user to maintain forecasts at any level within the aggregation and have these changes reflect both up and down the aggregation?
Promax PX can handle both Bottom-Up and Top-Down forecasting. Budgets and Strategy plans can be created at an aggregate level. All promotional uplift and baseline outputs are at the Account / SKU Level. Aggregations are handled by the system simply summing up the Account and SKU forecast.
Disaggregations are handled by the system using the statistically generated proportionate weightings of the Accounts and SKUs. These weightings can be derived from Scan Data, Sales Data or User Defined values. Users can define the desired date range that the system will use to determine the weightings.
Will Promax PX allow the entry of forecast using a variety of methods? The entry must be immediate, easy, and quick to perform to enable the forecaster to maintain their forecast in a timely fashion.
Promax PX facilitates all of these methods. Forecasts can be adjusted up and down by a percentage, manually entered, copied & pasted and baseline values can be imported from a flat file.